Journal of Heredity 2010:101(1):54–60 Ó The American Genetic Association. 2009. All rights reserved.

doi:10.1093/jhered/esp088 For permissions, please email: journals.permissions@oxfordjournals.org.

Advance Access publication December 4, 2009

A Candidate Gene Study of Canine Joint

Diseases

Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
DYLAN N. CLEMENTS, ANDREA D. SHORT, ANNETTE BARNES, LORNA J. KENNEDY, JOHN F. FERGUSON,
STEVEN J. BUTTERWORTH, NOEL FITZPATRICK, MATTHEW PEAD, DAVID BENNETT, JOHN F. INNES,
STUART D. CARTER, AND WILLIAM E. R. OLLIER
From Division of Veterinary Clinical Sciences, The Royal (Dick) School of Veterinary Studies, University of Edinburgh, Roslin
EH25 9RG, UK (Clements and Barnes); the Centre for Integrated Genomic Medical Research, The University of
Manchester, Manchester M13 9PT, UK (Clements, Short, and Ollier); the Musculoskeletal Research Group, School of
Veterinary Science, The University of Liverpool, Liverpool L69 7ZJ, UK (Kennedy, Innes, and Carter); the East Neuk
Veterinary Clinic, St Monans, Fife KY10 2DW, UK (Ferguson); the Weighbridge Referral Centre, Swansea SA6 8QF, UK
(Butterworth); the Fitzpartick Referrals, Farnham, Surrey GU10 2DZ, UK (Fitzpatrick); the Queen Mother Hospital for
Small Animals, The Royal Veterinary College, University of London, Hatfield AL9 7TA, UK (Pead); and the Division
Companion Animal Science, The University of Glasgow Veterinary School, Glasgow G61 1QH, UK (Bennett).

Address correspondence to Dylan N. Clements at the address above, or e-mail: dylan.clements@ed.ac.uk.

Abstract
Canine osteoarthritis (OA) commonly occurs in association with articular diseases, such as hip dysplasia (HD), elbow
dysplasia (ED), or cranial cruciate ligament rupture (CCLR). We hypothesized that a common genomic risk for the
development of canine joint disease and canine OA would be identified by evaluating the allele frequencies of candidate
gene single nucleotide polymorphisms (SNPs) in dogs with OA associated with different articular diseases when compared
with a general population of breed-matched dogs. DNA was extracted from blood samples obtained from Labrador
Retrievers and Golden Retrievers surgically treated for ED, HD, and CCLR and confirmed to have radiographic evidence of
OA. One hundred and thirteen SNPs in 20 candidate genes were genotyped. No significant associations were identified for
SNPs or haplotypes in the candidate genes for the diseases evaluated. The candidate gene approach for the study of genetic
association is unlikely to be successful for complex canine diseases such as OA without prior trait mapping evaluation.
Key words: association, canine, gene, joint, osteoarthritis

Osteoarthritis (OA) is a common, debilitating condition of
mammalian joints, characterized by the destruction of
articular cartilage, resulting in pain and dysfunction of the
affected joint. OA is estimated to affect up to 20% of dogs
over one year of age (Paradis et al. 2003) in the general dog
population. The joints most commonly affected by OA in the
dog are the hip, the elbow, and the stifle. Historically, OA of
these joints has been considered to be secondary to primary
diseases, such as hip dysplasia (HD), elbow dysplasia (ED),
and cranial cruciate ligament rupture (CCLR).
Breed risks for the development of HD, ED, and CCLR
are marked. The Labrador Retriever demonstrates a 3.4-fold
increase in risk for the development of HD (LaFond et al.
2002), a 20.5-fold increase in risk for developing the primary
component of ED (LaFond et al. 2002; fragmented coronoid
process [FCP]), and a 5.5-fold increase risk for developing
CCLR (Duval et al. 1999). Sex predisposition to the
development of each of theses diseases also exist, with male
dogs being a greater risk for developing hip OA (Hays et al.
2007) and ED FCP; Salg et al. 2006, and neuter status
conferring an increased risk for developing CCLR (Whitehair
et al. 1993; Duval et al. 1999). The estimates of heritability for
HD vary between 0.18 and 0.74 (Reed et al. 2000; Wood et al.
2000, 2002; Todhunter et al. 2003; Janutta et al. 2005),
estimates for ED vary between 0.10 and 0.77 (Guthrie 1989;
Guthrie and Pidduck 1990; Grondalen and Lingaas 1991;
Studdert et al. 1991; Maki et al. 2000), and estimates for CCLR
vary between 0.27 and 0.31 (Nielen et al. 2003; Wilke et al.
2006). A genetic correlation between HD and ED has also
been identified in a population of Rottweilers (Maki et al.
2000), suggesting that these traits may be influenced by the
same genetic and/or environmental factors in certain breeds.
As both ED (Lang et al. 1995) and HD (Wood et al. 2002) are
phenotyped by radiographic measures including osteophyto-
sis, the genetic risk factors which are common between these 2
diseases may code for a risk of developing osteophytosis.

54
 Clements et al.
A Candidate Gene Study of Canine Joint Diseases
Recent evidence suggests that genetic factors may
additionally affect the development of OA in dogs affected
by articular disease. Differences in the breed tolerance of
passive hip laxity for the development of hip OA imply that
genetic differences affect the development of canine hip OA
(Smith et al. 2001). The severity of OA in dogs presenting
with ED, HD, and CCLR can vary widely, which is
a function of disease duration, animal activity, nutrition
status, and genetic profile. Thus, although the significance of
primary versus secondary canine OA is undetermined,
canine OA per se is likely to have a significant genetic
background.
In contrast to human OA (Clements et al. 2006), the
genomic basis of joint disease in dogs has received limited
investigation to date. A previous case-control candidate
gene study of single nucleotide polymorphisms (SNPs) in 4
candidate genes (Fibronectin 1, type 9 collagen alpha 1
chain, type 9 collagen alpha 2 chain, and cartilage
oligometric protein) failed to identify any significant
associations between SNPs and the development of CCLR
in a population of Newfoundland dogs (Wilke et al. 2005). A
study of microsatellite markers adjacent to 14 candidate
collagen genes for ED also failed to identify significant
associations with the development of the disease (Salg et al.
2006). A microsatellite marker (FH2320) on canine
chromosome 3 (CFA3) has been linked with the de-
velopment of osteophytosis of the cranial and caudal
acetabulum in Portuguese water dogs (Chase et al. 2005). In
a separate study of a Labrador Retriever and Greyhound
cross pedigree, putative quantitative trait loci contributing to
macroscopic evidence of hip OA were identified on CFA05,
18, 23, and 31 (Mateescu et al. 2008). Further quantitative
assessment of the same pedigree also revealed that hip OA
was inherited additively and without dominance (Hays et al.
2007).
The majority of candidate genes studies in human OA
have evaluated genes that were associated with the
molecular pathogenesis of the disease, such as cytokines
and structural components of the extracellular matrix
(Clements et al. 2006). The most successful study in-
vestigating candidate gene associations with human OA
focused on genes that are differentially expressed in human
OA synovium and cartilage (Valdes et al. 2004). Sub-
sequently, a number of these gene polymorphism associa-
tions have been reproduced in different disease cohorts
(Valdes et al. 2007). Furthermore, a prediction of OA risk in
individuals could be attained by combining several of the
genes that were consistently shown to be involved in OA
susceptibility (Valdes, Doherty, et al. 2008).
We hypothesized that genomic risk for the development
of joint disease per se and canine OA would be similar
across the 3 most common articular diseases affecting dogs
(ED, HD, and CCLR) and between different breeds of dog
for the same disease (CCLR). We further hypothesized that
these genomic risks could be elucidated by evaluating the
allele frequencies of SNPs in candidate genes in populations
of dogs with ED, HD, and CCLR of a single breed and
between 2 breeds with a common disease (CCLR).
Materials and Methods
Candidate Gene Selection
Candidate genes were selected on the basis of previous
association of polymorphisms reported with OA in man or
from differential gene expression in articular tissues from
canine or human OA joints. A full list of the genes evaluated
the SNP positions, gene function, and justification for
Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
inclusion as a candidate gene are reported in Table 1.
The genes selected were ankyrin repeat domain 10
(ANKRD10), adenosine triphosphatase, Class VI, Type 11B
(ATP11B), interleukin 1 alpha (IL1a), -2 (IL2), -4 (IL4), -6
(IL6), -10 (IL10), -12b (IL12b), leptin receptor (LEPR),
matrix metalloproteinase 3 (MMP3), -9 (MMP9), -13
(MMP13), secreted protein, acidic, cysteine rich (osteonectin/
SPARC), tissue inhibitor of metalloproteinase 1 (TIMP1),
-2 (TIMP2), -3 (TIMP3), -4 (TIMP4), tenascin C (TNC),
tumor necrosis factor a (TNFa), Zinc finger, and SWIM-
type containing 2 (ZSWIM2).
Cohort Collection
Genomic DNA was extracted from residual clotted and
ethylenediaminetetraacetic acid (EDTA) preserved blood
samples using a standard phenol–chloroform extraction
method. Samples were suspended in Tris–EDTA and
normalized to 5 ng/ll. All case samples were obtained
from the United Kingdom DNA Archive for Companion
Animals (http://pcwww.liv.ac.uk/DNA_Archive_for_
Companion_Animals), and all control samples were
obtained from a population of breed-matched dogs from
the United Kingdom undergoing vaccination. Control dogs
were not phenotyped for disease. The breeds and
orthopedic diseases evaluated were selected by choosing
cohorts for which at least 30 samples had been collected
from individuals within a single breed, for a given
condition. All samples from cases were collected by
veterinary orthopedic specialist surgeons from dogs
surgically treated for ED, HD, or CCLR and with no
clinical evidence of a concurrent orthopedic condition
(ED, HD, or CCLR) at the time of treatment. All cases had
radiographic evidence of OA of the affected joint at the
time of surgery. Samples were collected from Labrador
Retrievers and Golden Retrievers being surgically treated
for ED (FCP; Labrador Retrievers, n 5 81 [11 females, 7
female neutered, 61 males, 1 male neutered, 1 male
neutering status unknown]), HD (Labrador Retrievers,
n 5 32 [9 females, 7 female neutered, 14 males, 1 male
neutered, 1 male unknown neutering status]), CCLR
(Labrador Retrievers n 5 51 [10 females, 19 female
neutered, 15 males, 7 male neutered], Golden Retrievers 45
[5 females, 26 female neutered, 10 males, 4 male neutered]),
and a population of dogs undergoing vaccination (Labrador
Retrievers n 5 344 [89 females, 29 female neutered, 105
males, 28 male neutered, 93 sex and neuter status
unknown], Golden Retrievers n 5 94 [42 females, 11
female neutered, 34 males, 6 male neutered, 1 sex and
neutered status unknown]). The control samples were not
55
Journal of Heredity 2010:101(1)

Table 1. Candidate genes selected for evaluation in canine OA and their SNP positions

Justification for
Name CFA U P E I D Function evaluation inclusion References
Ankyrin repeat 22 2 3 Structural Increased expression in (Clements et al. 2008)
domain 10 component of canine ruptured CCL
(ANKRD10) muscle
Adenosine 34 6 Cell membrane Increased expression in (Clements et al. 2008)
triphosphatase, transport canine ruptured CCL

Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
Class VI, Type 11B
(ATP11B)
Interleukin 1 alpha 17 3 2 2 Proinflammatory Increased expression in (Loughlin et al. 2002;
(IL1a) cytokine canine OA synovium Smith et al. 2004, 2005;
and ruptured CCL. SNPs Maccoux et al. 2007)
associated with human OA
Interleukin 10 (IL10) 7 3 2 3 3 2 Antiinflammatory Increased expression in canine (Maccoux et al. 2007)
cytokine OA synovium
Interleukin 12 (IL12B) 4 5 4 2 Proinflammatory Increased expression in human (Sakkas et al. 1998)
cytokine OA synovium
Interleukin 2 (IL2) 19 1 T and B cell Expressed in human OA (Kardel et al. 2003)
proliferation synovium
Interleukin 4 (IL4) 11 1 1 4 2 Antiinflammatory Demonstrates role in sexual (Mahr et al. 2003)
cytokine dimorphisms OA susceptibility
in experimental animal model
Interleukin 6 (IL6) 14 2 2 1 3 Proinflammatory Expressed in canine OA (Maccoux et al. 2007)
cytokine synovium and ruptured CCL
Leptin receptor 5 3 Adipokine receptor Increased agonist (leptin) (Dumond et al. 2003)
(LEPR) expression in human OA
cartilage
Matrix metalloproteinase 5 2 6 Collagenase (cartilage Increased expression in canine hip (Clements et al. 2006b,
13 (MMP13) break down) OA cartilage and ruptured CCL 2008)
Matrix metalloproteinase 5 1 1 Collagenase (cartilage Increased expression in canine (Hegemann et al. 2003)
3 (MMP3) break down) stifle OA
Matrix metalloproteinase 24 2 1 1 Gelatinase Increased expression in canine hip (Clements et al. 2006b,
9 (MMP9) OA cartilage and ruptured CCL 2008)
Secreted protein, acidic, 4 4 3 Matrix-associated Increased expression in canine (Clements et al. 2008)
cysteine rich protein ruptured CCL
(osteonectin/SPARC)
Tissue inhibitor of X 1 1 Inhibition of Increased expression in canine hip (Clements et al. 2006b)
metalloproteinase 1 metalloproteinase OA cartilage
(TIMP1) activity
Tissue inhibitor of 9 1 Inhibition of Reduced expression in canine hip (Clements et al. 2006b,
metalloproteinase 2 metalloproteinase OA cartilage and ruptured CCL 2008)
(TIMP2) activity
Tissue inhibitor of 10 1 1 Inhibition of Increased expression in human (Kevorkian et al. 2004)
metalloproteinase 3 metalloproteinase OA cartilage
(TIMP3) activity
Tissue inhibitor of 20 2 2 Inhibition of Reduced expression in canine hip (Clements et al. 2006b)
metalloproteinase 4 metalloproteinase cartilage OA
(TIMP4) activity
Tenascin C (TNC) 11 3 2 2 1 Extracellular matrix Increased expression in ruptured (Salter 1993; Clements
protein CCL/OA cartilage et al. 2006b, 2008)
Tumor necrosis factor 12 2 1 1 3 2 Proinflammatory Expressed by canine OA (Fujita et al. 2005;
alpha (TNFa) cytokine synovium Hegemann et al. 2005)
Zinc finger, SWIM-type 36 2 2 Metabolism Increased expression in OA (Clements et al. 2008)
containing 2 (ZSWIM2) cartilage and ligament

CFA, canine chromosome; U, upstream; P, Promotor; E, Exon; I, Intron; D, Downstream and their justification for evaluation as candidate genes.

evaluated for disease status. An internal genotyping control
was included on each plate.
SNP Identification
Selected regions of each candidate gene were amplified by
polymerase chain reaction (PCR). The PCR product was
assessed for the presence of a polymorphic product using
denaturing high-performance liquid chromatography
(Spiegelman et al. 2000), and amplicons with melt curve
analyses indicating a SNP were sequenced. SNPs identified
were annotated to a genomic position by alignment of the
sequence with the canine genome (www.Ensembl.org), and

56
 Clements et al.
A Candidate Gene Study of Canine Joint Diseases
the Ensembl gene annotations (exon numbers, size, and
position) were used to define the SNP positions. SNPs were
designated a genomic location on the basis of their position
(pre-gene  10 000-bp upstream of exon 1, promotor
,1000 bp of the start of exon 1, post-gene  10 000-bp
downstream of the last exon). Forty-six SNPs were selected
from a previous study (IL1a, IL2, IL4, IL6, IL10, IL12b,
and TNFa; Short 2006; Short et al. 2009), 24 SNPs were
identified as described (ATP11B, ANKRD10, LEPR,
MMP13, SPARC, TNC, ZSWIM2). A further 43 SNPs
were selected from the published canine genome sequence
(www.ncbi.nlm.nih.gov) and the Broad Institute canine
SNP database (http://www.broad.mit.edu/tools/data/
genvar.html).
Genotyping
Genotyping was performed using the Sequenom MASSarray
platform (Sequenom, Hamburg, Germany) as previously
described (Short et al. 2007). Briefly, primers and probes
were designed using the Sequenom assay design software
version 3 and synthesized by Metabion AG (Martinsried,
Germany). Primers and probes were pooled as recom-
mended by the manufacturers instructions (http://www.
sequenom.com). Multiplex PCR reactions, product cleanup,
and probe extension reaction were performed in 384-well
plates with 20 ng of DNA per well, using iPLEX Gold
reagents. Samples were diluted and desalted with 6 mg of
resin before dispensation onto a SpectroCHIP (Sequenom)
using the Sequenom nanodispenser, before genotype identi-
fication by matrix-assisted laser desorption/ionization–time
of flight mass spectrometry.
Data Analysis
Genotype and phenotype data were imported into BCgene
software (www.bcplatforms.com), which was used to
calculate genotyping rates, minor allele frequencies (MAFs),
and Hardy–Weinberg equilibrium (HWE) for each control
population. SNPs or individuals were not analyzed further if
the call rates were below 80% or if the control population
was not in HWE. Analysis of genotype associations with
disease was performed by logistical regression to correct for
the confounding variables (sex and neuter status) using
PLINK, a web-based genetic analysis program (Purcell et al.
2007), with additive, dominant, and recessive models. The
strength of the significant associations was checked with
a multiple permutation test, using 10 000 permutations.
Haplotype frequencies were calculated for each gene, and
haplotype associations with disease were performed using
logistical regression to correct for the confounding variables
using PLINK. The strength of the associations was again
checked with a multiple permutation test, using 10 000
permutations.
Estimates of statistical power were performed using the
PS Power and Sample Size program (Dupont and Plummer
1998). When SNPs with a MAF ranging from 5% to 50%
are used this study was powered to detected risk alleles with
Odds Ratios (OR) ranges of 1.64 –2.43 (Labrador Retrievers,
ED) to 2.10–3.46 (Golden Retrievers, CCLR) and protective
alleles with OR ranges of 0.61–0.13 (Labrador Retrievers,
ED) to 0.48—undetectable (Golden Retrievers, CCLR) if
the allele is protective, at 80% power (P , 0.05).
Results
Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
A total of 113 SNPs were analyzed; in Labrador Retrievers,
44% (n 5 54) of SNPs were informative (MAF .1%), 14%
(n 5 16) of SNPs demonstrated low heterozygosity (MAF
,1%), 15% (n 5 17) were monoallelic, and 27% (n 5 30)
were not in HWE; in Golden Retrievers, 61% (n 5 69) were
informative, 27% (n 5 31) were monoallelic, and 12% (n 5
13) were not in HWE. The average genotyping rate for the
Labrador Retrievers was 94.5% (range 81.3–99.6%), and
96.6% for the Golden Retrievers samples (range 84.9–
100%). The concordance of the internal genotyping control
between plates was 100%.
Case-control comparison of genotypes identified 10
SNPs whose minor alleles were significantly associated with
risk (n 5 5) or protection (n 5 5) of orthopedic disease
independent of confounding using the additive model
(Table 2). The case-control comparison of haplotype
frequencies identified 8 haplotypes that were associated
with the risk (n 5 5) or protection (n 5 3) of orthopedic
disease independent of confounding using the additive
model (Table 3). One haplotype of IL12b (CCTAACGG)
was associated with the risk of developing ED and HD in
Labrador Retrievers. Multiple permutation tests revealed
statistical significance to be lost for all SNPs and haplotypes
evaluated.
Discussion
A common genomic risk for joint disease and OA was not
identified between 2 different breeds for the same disease or
within the same breed for different diseases. Thus, our
hypothesis that the genomic basis for the development of
canine joint disease would be identified was not supported,
although many other plausible candidate genes were not
evaluated. Successful candidate gene studies of association
identifying positive associations between a canine phenotype
and genotype have been reported for diseases, such as
diabetes mellitus (Short 2006). However, the overall success
of such studies in identifying a mutation associated with
a trait is extremely low in dogs when compared with linkage
analysis (Aguirre-Hernandez and Sargan 2005). Although
the collection of data for mapping studies is challenging
from a client-owned dog population, whole-genome
association have the power to identify the genomic regions
which are associated with complex traits in dogs and thus
which harbor candidate genes suitable for further evaluation
(Karlsson and Lindblad-Toh 2008).
The analysis of multiple SNPs or haplotypes for
association with a phenotype requires appropriate statistical
correction to minimize the Type-1 statistical error (Balding
57
Journal of Heredity 2010:101(1)

Table 2. The significant susceptibility SNP associations

Position
Disease SNP SNP (CFA: base Minor MAF Major Corrected
(breed) Gene identity position pair) allele control allele OR (þ95% CI) P value P value
CCLR (GR) ANKRD10 ANK_I2 Intron 2-3 22:62038934 A 0.47 G 0.31 (0.13–0.75) 0.009 0.089
CCLR (GR) ANKRD10 rs23039786 Intron 1-2 22:62043702 G 0.46 A 0.33 (0.14–0.77) 0.011 0.123
CCLR (GR) TNC TNC_P4 Promotor 11:72171896 T 0.25 C 2.65 (1.25–5.61) 0.011 0.146
CCLR (GR) TNC TNC_E2 Exon 2 11:72170870 G 0.43 A 1.89 (1.01–3.54) 0.046 0.588

Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
CCLR (GR) TNC rs22191687 Promotor 11:72171378 G 0.43 T 1.89 (1.01–3.54) 0.046 0.521
CCLR (GR) IL10 rs24430111 Post-gene 7:8886928 A 0.39 G 0.32 (0.1–0.95) 0.041 0.480
ED (LR) IL12b 12B_02W232 Pre-gene 4:54385156 A 0.15 T 0.51 (0.27–0.95) 0.033 0.594
ED (LR) IL12b 12B_02Y190 Pre-gene 4:54385114 T 0.14 C 0.50 (0.26–0.95) 0.034 0.601
HD (LR) IL4 4_13S97 Intron 2-3 11:23983033 G 0.18 C 2.04 (1.15–3.61) 0.015 0.298
HD (LR) IL4 4_8R458 Intron 3-4 11:23984915 A 0.19 G 2.02 (1.14–3.58) 0.016 0.319

SNP’s separated by breed (GR, Golden Retriever; LR, Labrador Retriever), disease (CCLR, cruciate disease; ED; HD), gene, SNP position, minor allele,
minor frequency in controls (%C), major allele, the OR for the MAFs in the disease population with 95% confidence intervals (95% CIs) and statistical
significance (P value) when compared by logistical regression, and the corrected P value determined by multiple permutation testing.

2006). The allele frequency of a number SNPs and
haplotypes were observed to be associated with different
joint disease phenotypes before correction by multiple
permutation testing. In most cases, the ORs for SNPs
associated with diseases evaluated in this study were below
the limits calculated for 80% statistical power, and these
limits were further reduced by the stratification of
populations on the basis of confounding variables.
A proportion of our control cases were likely to have
developed HD, ED, or CCLR with time, which would
artificially lower the differences and significance of the
changes in the MAFs recorded. Thus, the true ORs for the
MAFs would probably be greater if control populations
were available which had been accurately phenotyped for all
3 diseases. The true prevalence of each of these diseases in
the control populations in the United Kingdom are
unknown, so estimations of the degree of misclassification
in the control population could not be made, as robust
epidemiological studies of orthopedic diseases have not
been performed in dogs to date. Age was not included in the
statistical model, but sex and neuter status were in view of
previous associations with these factors with the diseases
evaluated.
Conversely, it should be noted that the phenotype quality
of the disease cohorts is extremely high, as all cases were at
the extreme end of the phenotype. First, each case required
surgery for their underlying condition and second, each case
was diagnosed by a veterinary orthopedic specialist.
Difficulties in obtaining appropriate numbers of samples
to accurately determine differences in allele and genotype
frequencies for case-control study of canine disease with
dog populations are a recognized issue (Short et al. 2007).
The majority of SNPs we evaluated were in the intronic
(n 5 33) or promotor (n 5 28) regions, in contrast to the
predominantly exonic SNPs evaluated in successful human
candidate gene association studies of OA (Valdes et al. 2004;

Table 3. The significant susceptibility haplotype associations

Haplotype
Number of frequency Corrected
Disease (breed) Gene haplotypes Haplotype (control) OR P value P value
CCLR (GR) ANKRD10 4 AGGG 0.54 2.53 0.010 0.184
CCLR (GR) ANKRD10 4 GAGG 0.24 0.42 0.031 0.533
CCLR (GR) TNC 5 CGGTTA 0.08 2.86 0.032 0.547
ED (LR) IL12b 4 CCTAACGG 0.16 1.64 0.042 0.686
ED (LR) IL12b 4 ATCTCTAG 0.14 0.52 0.045 0.710
ED (LR) TIMP3 3 AT 0.36 0.65 0.049 0.739
HD (LR) IL4 3 CCAGAG 0.19 2.06 0.014 0.273
CCLR (LR) IL12b 4 CCTAACGG 0.17 1.98 0.021 0.394

Haplotypes separated by breed (GR, Golden Retriever, LR, Labrador Retriever), disease (CCLR, cruciate disease; ED; HD), gene, haplotype frequency in
the control population, the OR for the haplotype frequency in the disease population, statistical significance (P value) when compared by logistical
regression, and the correct and the corrected P value determined by multiple permutation testing. The SNP identity (chromosome: base pair) used for
haplotypes for each gene were as follows: ANKRD10 (rs23039786 [22:62043702], ANK_I2 [22:62038934], ANK_I4 [22:62021737], rs23105168
[22:62018474]), TNC (TNC_P4 [11:72171896], rs22191687 [11:72171378], TNC_E2 [11:72170870], TNC_I4 [11:72165326], TNC_I24 [11:72112527],
TNC_E25 [11:72112377]), IL12B (12B_01M115 [4:54384525], 12B_01Y90 [4:54384501], 12B_02Y190 [4:54385114], 12B_02W232 [4:54385156],
12B_02Y407 [4:54385331], 12B_03Y82 [4:54385528], 12B_03R196 [4:54385642], 12B_03R462 [4:54385908]), TIMP3 (rs22007793 [10:33774314],
rs22007745 [10:33749146]), IL4 (rs22146864 [11:23976639], 4_13S97 [11:23983033], 4_12M397 [11:23982900], 4_8R458 [11:23984915], 4_2M351
[11:23987559], rs22189535 [11:23988181]).

58
 Clements et al.
A Candidate Gene Study of Canine Joint Diseases
Valdes, Van-Oene, et al. 2006). Ideally, exonic SNPs would
have been evaluated for all genes, but few or none were
identified in the genes screened. Although open-access
canine SNP databases are available, detail is presently
lacking for individual genes and was not always able to
detect exonic SNPs in each candidate gene evaluated using
our methodology.
Conclusions
A common genetic risk for the development of canine joint
disease and OA was not identified between different dis-
eases within a single breed, or for the same disease in two
different breeds in this study of association with poly-
morphisms in 20 candidate genes. Spurious association of
candidate gene allele frequencies is ready identified in can-
didate genes if data are not appropriately corrected for
multiple testing.
Funding
PetPlan Charitable Trust and the Biotechnology and
Biological Sciences Research Council.
References
Aguirre-Hernandez J, Sargan DR. 2005. Evaluation of candidate genes in
the absence of positional information: a poor bet on a blind dog!. J Hered.
96:475–484.
Balding DJ. 2006. A tutorial on statistical methods for population
association studies. Nat Rev Genet. 7:781–791.
Chase K, Lawler DF, Carrier DR, Lark KG. 2005. Genetic regulation of
osteoarthritis: a QTL regulating cranial and caudal acetabular osteophyte
formation in the hip joint of the dog (Canis familiaris). Am J Med Genet.
135:334–335.
Clements DN, Carter SD, Innes JF, Ollier WE. 2006a. Genetic basis of
secondary osteoarthritis in dogs with joint dysplasia. Am J Vet Res.
67:909–918.
Clements DN, Carter SD, Innes JF, Ollier WE, Day PJ. 2006b. Analysis of
normal and osteoarthritic canine cartilage mRNA expression by quantitative-
PCR. Arthritis Res Ther. 8:R158.
Clements DN, Carter SD, Innes JF, Ollier WER, Day PJ. 2008. Gene
expression profiling of normal and ruptured canine anterior cruciate
ligaments. Osteoarthritis Cartilage. 16:195–203.
Dumond H, Presle N, Terlain B, Mainard D, Loeuille D, Netter P, Pottie P.
2003. Evidence for a key role of leptin in osteoarthritis. Arthritis Rheum.
48:3118–3129.
Dupont WD, Plummer WD. 1998. Power and sample size calculations for
studies involving linear regression. Control Clin Trials. 19:589–601.
Duval JM, Budsberg SC, Flo GL, Sammarco JL. 1999. Breed, sex, and body
weight as risk factors for rupture of the cranial cruciate ligament in young
dogs. J Am Vet Med Assoc. 215:811–814.
Fujita Y, Hara Y, Nezu Y, Yamaguchi S, Schulz KS, Tagawa M. 2005.
Direct and indirect markers of cartilage metabolism in synovial fluid
obtained from dogs with hip dysplasia and correlation with clinical and
radiographic variables. Am J Vet Res. 66:2028–2033.
Grondalen J, Lingaas F. 1991. Arthrosis in the elbow joint of young rapidly
growing dogs—a genetic investigation. J Small Anim Pract. 32:460–464.
Guthrie S. 1989. Some radiographic and clinical aspects of ununited
anconeal process. Vet Rec. 124:661–662.
Guthrie S, Pidduck HG. 1990. Heritability of elbow osteoachondrosis
within a closed population of dogs. J Small Anim Pract. 31:93–96.
Hays L, Zhang Z, Mateescu RG, Lust G, Burton-Wurster NI, Todhunter
RJ. 2007. Quantitative genetics of secondary hip joint osteoarthritis in
Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
a Labrador Retriever-Greyhound pedigree. Am J Vet Res. 68:35–41.
Hegemann N, Wondimu A, Kohn B, Brunnberg L, Schmidt MF. 2005.
Cytokine profile in canine immune-mediated polyarthritis and osteoarthritis.
Vet Comp Orth Trauma. 18:67–72.
Hegemann N, Wondimu A, Ullrich K, Schmidt MFG. 2003. Synovial
MMP-3 and TIMP-1 levels and their correlation with cytokine expression in
canine rheumatoid arthritis. Vet Immunol Immunopathol. 91:199–204.
Janutta V, Hamann H, Klein S, Tellhelm B, Distl O. 2005. Genetic
evaluation of elbow angles as predictors of elbow dysplasia in German
shepherd dogs. J Vet Med A Physiol Pathol Clin Med. 52:254–261.
Kardel R, Ulfgren A-K, Reinholt FP, Holmlund A. 2003. Inflammatory cell
and cytokine patterns in patients with painful clicking and osteoarthritis in
the temporomandibular joint. Int J Oral Maxillofac Surg. 32:390–396.
Karlsson EK, Lindblad-Toh K. 2008. Leader of the pack: gene mapping in
dogs and other model organisms. Nat Rev Genet. 9:713–725.
Kevorkian L, Young DA, Darrah C, Donell ST, Shepstone L, Porter S,
Brockbank SM, Edwards DR, Parker AE, Clark IM. 2004. Expression
profiling of metalloproteinases and their inhibitors in cartilage. Arthritis
Rheum. 50:131–141.
LaFond E, Breur GJ, Austin CC. 2002. Breed susceptibility for
developmental orthopedic diseases in dogs. J Am Anim Hosp Assoc.
38:467–477.
Lang J, Busato A, Baumgartner D, Fluckiger M. 1995. Comparison of
two classification protocols in the evaluation of elbow disease in the
Bernese mountain dog. In: Abstracts of Papers Presented at the Annual
Meeting of the International Elbow Working Group (IEWG). 1994
August 7; Philadelphia, PA. Veterinary Radiology & Ultrasound. 36. p. 170–
171.
Loughlin J, Dowling B, Mustafa Z, Chapman K. 2002. Association of the
interleukin-1 gene cluster on chromosome 2q13 with knee osteoarthritis.
Arthritis Rheum. 46:1519–1527.
Maccoux LJ, Salway F, Day PJR, Clements DN. 2007. Expression profiling
of select cytokines in canine osteoarthritis tissues. Vet Immunol
Immunopathol. 118:59–67.
Mahr S, Menard J, Krenn V, Muller B. 2003. Sexual dimorphism in the
osteoarthritis of STR/ort mice may be linked to articular cytokines. Ann
Rheum Dis. 62:1234–1237.
Maki K, Liinamo AE, Ojala M. 2000. Estimates of genetic parameters for
hip and elbow dysplasia in Finnish Rottweilers. J Anim Sci. 78:
1141–1148.
Mateescu RG, Burton-Wurster NI, Tsai K, Phavaphutanon J, Zhang Z,
Murphy KE, Lust G, Todhunter RJ. 2008. Identification of quantitative trait
loci for osteoarthritis of hip joints in dogs. Am J Vet Res. 69:1294–1300.
Nielen AL, Knol BW, van Hagen MA, van-der-Gaag I. 2003. Genetic and
epidemiological investigation of a birth cohort of boxers. Tijdschr
Diergeneeskd. 128:586–590.
Paradis M, Sauve F, Charest J, Refsal KR, Moreau M, Dupuis J. 2003.
Effects of moderate to severe osteoarthritis on canine thyroid function. Can
Vet J. 44:407–412.
Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D,
Maller J, Sklar P, de Bakker PIW, Daly MJ, et al. 2007. PLINK: a tool set
59
Journal of Heredity 2010:101(1)
for whole-genome association and population-based linkage analyses. Am
J Hum Genet. 81:559–575.
Reed AL, Keller GG, Vogt DW, Ellersieck MR, Corley EA. 2000. Effect of
dam and sire qualitative hip conformation scores on progeny hip
conformation. J Am Vet Med Assoc. 217:675–680.
Sakkas LI, Johanson NA, Scanzello CR, Platsoucas CD. 1998. Interleukin-
12 is expressed by infiltrating macrophages and synovial lining cells in
rheumatoid arthritis and osteoarthritis. Cell Immunol. 188:105–110.
Salg KG, Temwitchitr J, Imholz S, Hazewinkel HA, Leegwater PA. 2006.
Assessment of collagen genes involved in fragmented medial coronoid
process development in Labrador Retrievers as determined by affected
sibling-pair analysis. Am J Vet Res. 67:1713–1718.
Salter DM. 1993. Tenascin is increased in cartilage and synovium from
arthritic knees. Br J Rheumatol. 32:780–786.
Short AD. 2006. The genetics of canine diabetes: a candidate gene study
[PhD thesis]. [Manchester (UK)]: The University of Manchester.
Short AD, Catchpole B, Kennedy LJ, Barnes A, Fretwell N, Jones C,
Thomson W, Ollier WER. 2007. Analysis of candidate susceptibility genes
in canine diabetes. J Hered. 98:518–525.
Short AD, Catchpole B, Kennedy LJ, Barnes A, Lee AC, Jones CA,
Fretwell N, Ollier WER. 2009. T cell cytokine gene polymorphisms in
canine diabetes mellitus. Vet Immunol Immunopathol. 128:137–146.
Smith AJ, Elson CJ, Perry MJ, Bidwell JL. 2005. Accuracy of haplotype
association studies is enhanced by increasing number of polymorphic loci
examined: comment on the article by Meulenbelt et al. Arthritis Rheum.
52:675–676.
Smith AJ, Keen LJ, Billingham MJ, Perry MJ, Elson CJ, Kirwan JR, Sims
JE, Doherty M, Spector TD, Bidwell JL. 2004. Extended haplotypes and
linkage disequilibrium in the IL1R1-IL1A-IL1B-IL1RN gene cluster:
association with knee osteoarthritis. Genes Immun. 5:451–460.
Smith GK, Mayhew PD, Kapatkin AS, McKelvie PJ, Shofer FS, Gregor TP.
2001. Evaluation of risk factors for degenerative joint disease associated
with hip dysplasia in German shepherd dogs, golden retrievers, Labrador
retrievers, and Rottweilers. J Am Vet Med Assoc. 219:1719–1724.
Spiegelman JI, Mindrinos MN, Oefner PJ. 2000. High-accuracy DNA
sequence variation screening by DHPLC. Biotechniques. 29:1084–1090,
1092.
Studdert VP, Lavelle RB, Beilharz RG, Mason TA. 1991. Clinical-features
and heritability of osteochodrosis of the elbow in Labrador retrievers.
J Small Anim Pract. 32:557–563.
60
Todhunter RJ, Bliss SP, Casella G, Wu R, Lust G, Burton-Wurster NI,
Williams AJ, Gilbert RO, Acland GM. 2003. Genetic structure of
susceptibility traits for hip dysplasia and microsatellite informativeness of
an outcrossed canine pedigree. J Hered. 94:39–48.
Valdes AM, Hart DJ, Jones KA, Surdulescu G, Swarbrick P, Doyle DV,
Schafer AJ, Spector TD. 2004. Association study of candidate genes for the
prevalence and progression of knee osteoarthritis. Arthritis Rheum.
50:2497–2507.
Downloaded from http://jhered.oxfordjournals.org/ at Fundação Coordenação de Aperfeiçoamento de Pessoal de NÃ-vel Superior on June 1, 2014
Valdes AM, Loughlin J, Oene MV, Chapman K, Surdulescu GL, Doherty
M, Spector TD. 2007. Sex and ethnic differences in the association of
ASPN, CALM1, COL2A1, COMP, and FRZB with genetic susceptibility to
osteoarthritis of the knee. Arthritis Rheum. 56:137–146.
Valdes AM, Doherty M, Spector TD. 2008. The additive effect of individual
genes in predicting risk of knee osteoarthritis. Ann Rheum Dis. 67:124–127.
Valdes AM, Van-Oene M, Hart DJ, Surdulescu GL, Loughlin J, Doherty M,
Spector TD. 2006. Reproducible genetic associations between candidate
genes and clinical knee osteoarthritis in men and women. Arthritis Rheum.
54:533–539.
Whitehair JG, Vasseur PB, Willits NH. 1993. Epidemiology of cranial
cruciate ligament rupture in dogs. J Am Vet Med Assoc. 203:
1016–1019.
Wilke VL, Conzemius MC, Rothschild MF. 2005. SNP detection and
association analyses of candidate genes for rupture of the cranial cruciate
ligament in the dog. Anim Genet. 36:519–521.
Wilke VL, Conzemius MG, Kinghorn BP, Mcrossan PE, Weiguo C,
Rothschild MF. 2006. Inheritance of rupture of the cranial cruciate ligament
in Newfoundlands. J Am Vet Med Assoc. 228:61–64.
Wood JLN, Lakhani KH, Dennis R. 2000. Heritability and epidemiology of
canine hip-dysplasia score in flat-coated retrievers and Newfoundlands in
the United Kingdom. Prev Vet Med. 46:75–86.
Wood JLN, Lakhani KH, Rogers K. 2002. Heritability and epidemiology of
canine hip-dysplasia score and its components in Labrador retrievers in the
United Kingdom. Prev Vet Med. 55:95–108.
Received December 24, 2008; Revised August 29, 2009;
Accepted September 25, 2009
Corresponding Editor: Elaine Ostrander